2.0 Analysis 2.1 Introduction The navigation officers had not been made aware of the proposed manoeuvres for departure. The master adopted an independent command role and there was no discussion of how the departure from the berth was to be accomplished. Without the back-up, monitoring and support of a senior experienced officer and some division of responsibility, the chances of a successful departure were reduced. The conclusions of the analysis of the physical evidence do not support the version of the sequence of events obtained from interviews with the ship's personnel. 2.2 Damage to the Vessel An underwater video survey of the damage sustained by the vessel showed that all the damage appeared to run in a fore-and-aft line, on either side of the keel, from the stem to the collision bulkhead. The forefoot was lifted extensively and the damage ran in a fore-and-aft direction, indicating that the damage resulted from a high impact grounding from directly ahead while the vessel was moving ahead. The type and extent of the damage incurred by the vessel does not support the master's testimony that the vessel struck a pinnacle while moving ahead, drifted ashore and grounded laterally. 2.3 Reported Underwater Obstruction Because the master indicated that there was an uncharted pinnacle in the area of the grounding position, sweeps of the area were carried out. No obstruction was found. Given that the sweeps carried out by the CCG were directed to the reported position of the obstruction and that the area had been surveyed by the Canadian Hydrographic Service in May 1993, it is unlikely that such an obstruction exists. Further, an examination of the video film of the sea bottom and the kelp in the area where the vessel was hard aground did not indicate that the vessel had drifted laterally to her grounded position. The vessel's echo-sounder trace paper was examined but there was no evidence of a trace. The report that the digital depth read-out was in operation at the time of the grounding cannot be verified. It is not known why the depth alarm did not sound if it was set as reported. 2.4 Decision to Sail at Night As he had obtained information from knowledgeable persons to assist him in leaving the berth, the master believed that he did not require to engage a pilot for departure, contrary to what he had been instructed to do. It is evident that the master's local knowledge in the unfamiliar port was insufficient to sail at night. He was under no pressure to do so. The chances of an unassisted successful departure would have been increased by waiting for daylight. 2.5 Reconstruction of the Departure By referring to the engine movements as recorded in the bell book, to the course-recorder trace and to the vessel's manoeuvring data (this information is tabulated in Appendix C), it was possible to reconstruct the approximate movements of the vessel from the berth to the grounding position. It can be seen from the sketch of the area (see Appendix B) that, at about 0215, the vessel crossed the 15 m contour when vibration was felt. The vessel would have had an UKC of less than 1 m when making the starboard swing to clear Pig Point with engines at full ahead. In the shallow water, the vessel's manoeuvrability would have been hampered by squat. As the speed over the ground increased, the vessel's squat would have decreased the UKC and the swing to starboard would have become more laboured due to these hydrodynamic effects. This would also have increased the vessel's advance and transfer, and decreased her turning efficiency. It is likely that the master became aware of this situation because speed was reduced at 0216. Attempts were made to tighten the vessel's turning circle by the use of the bow thruster at 0217, but the vessel was, at this time, moving ahead at a speed of two to three knots, and the effect of the bow thruster would have been minimal. The engines were not put astern to reduce speed and squat, and to increase bow thruster efficiency. In fact, the engines were not put astern until after the grounding. Speed was again increased to full ahead at 0218 and at 0220, immediately before the vessel went aground. Between these times, however, the vessel's UKC and manoeuvrability were further decreasing in the rapidly shoaling water. Given the manoeuvres carried out from the time that the vessel crossed the 15 m contour, the vessel's grounding was inevitable. 2.6 Departure Procedure Given that the vessel had been berthed port side to the dock and considering the physical limitations of the cove, two alternate departure strategies were available to the master. One alternative was to swing the vessel short-round off the berth and proceed out of the bay on the south-westerly heading which would be taken by a vessel which had been conventionally secured starboard side to the dock. The second, which the master elected to attempt, was to take the vessel on a southerly heading between the shoal and Pigeon Head, after making a tight turn to starboard immediately on clearing the berth. Either method is feasible if the hazards inherent in the procedure are recognized. There was adequate water depth near the berth for either manoeuvre to be carried out, and the required turn can be initiated by the use of the vessel's moorings, with or without the assistance of a tug. Once the vessel's bow or stern has cleared the dock, a swing can be induced by alternating the main engine ahead and astern while applying the appropriate helm. The bow thruster can also be used to maximum advantage to assist when the vessel is not making way through the water. The first manoeuvre required that the vessel swing through 180 and the second that the vessel swing through some 90. In the present case, the system of parallel indexing employed by the master did not alert him to the fact that the vessel had made too much headway before completing the required 90 change in heading. To be effective, the parallel indexing technique must incorporate a means of monitoring the vessel's progress. Also, in this instance, the capability of the ARPA to provide sophisticated navigational support was not used to supplement the system of parallel indexing employed. 2.7 Visibility from the Bridge After all the mooring lines had been let go, the crew members were engaged in clewing up for sea which necessitated the use of the deck working lights. Although it was deemed necessary for this operation, the deck lighting was detrimental to visibility from the bridge. Visibility from the bridge was not reported to have been a factor. People on the bridge would not have been able to obtain full night vision. This would also have caused a reduction in situational awareness and of how the vessel was responding to the master's manoeuvres. Because no look-out was posted on the forecastle head, clear of the deck lighting, there was no possibility of an early warning that the vessel was closing on the land. 3.0 Conclusions 3.1 Findings The master decided to sail at night from a port with which he was unfamiliar. The master did not engage the services of either the pilot or the tug available to assist him. The vessel was not turned short-round in the area of the berth to enable her to follow the recognized departure procedure. The master's departure plan was known only to him; he did not discuss it with the rest of the bridge team. There was no senior officer on the bridge to assist the master during departure, and the officer of the watch (OOW) was assigned too many duties to enable him to assist the master or monitor the vessel's progress. The master's departure plan was not laid down on the large-scale survey field sheets of the area which had been provided to him before departure. The vessel's working lights were not turned off for night-time departure nor was a look- out posted where these lights would not hamper visibility. The only conventional navigational aid in the area, an isolated danger buoy marking the shoal to the south of the berth, is seasonal and had been removed for the winter. The master relied solely on parallel indexing for navigation despite the presence of various conspicuous radar targets. All engine movements before the grounding were in the ahead mode. A video film of the vessel aground and of the adjacent sea-bed indicate that the vessel struck the bottom head-on and remained aground; she did not drift laterally to her grounded position. The master's testimony and the entries relative to the grounding made in official documents are inconsistent with the visible damage sustained by the vessel. When the bow thruster compartment flooded, the breakers of the vessel's service generators tripped because the bow thruster breaker had a trip delay exceeding that of the generator breakers. The emergency generator did not start automatically after the electrical black-out because it was not switched to the automatic start position. 3.2 Causes The CSL ATLAS grounded because the master sailed at night from a port with which he was unfamiliar, did not employ the recognized departure procedure, did not establish either a bridge resource management regime or a voyage plan for leaving the berth, and did not engage either the pilot or tug available. 4.0 Safety Action 4.1 Action Taken 4.1.1 Bridge Resource Management (BRM) The owners of the CSL ATLAS, CSL International (CSL), have engaged the Centre for Marine Simulation in St. John's, Newfoundland, to develop a course and train their ships' officers in effective Bridge Resource Management (BRM) techniques. The company has indicated that all CSL masters and chief officers will attend the BRM course during 1995. 4.1.2 Passage Planning To ensure that masters and navigation officers have the necessary skills to effectively develop passage plans for berth-to-berth navigation, CSL has developed a Passage Planning and Navigation Checklist. The procedural documents are submitted to management for audit on a voyage-by-voyage basis. Further, to improve proficiency in port manoeuvre planning, CSL has also set up a training program for its masters and navigation personnel.